Issue was in fact caused by wrong DNA storage, which
apparently was considering unsigned long as just 4
bytes here.
Now use uint64_t to be sure timestamp does fit into
storage on all the platforms.
Thanks Campbell for help looking in the issue.
GE was copying the pose channels without increasing user counter for a
custom bone shape object. Freeing copied pose will for give decrement
errors.
The same increment issue seems to happen in BKE_pose_copy_data, which
is also solved now.
Image empties don't actually support mipmaps right now, but the user preference
doubles as a way to disable any kind of interpolating in texture filter and just
show nearest neighbour filter, so for this is a bit more consistent.
Issue was caused by wm->defaultconf being NULL when in
background mode which made keymap modifications from a
script crash.
Reviewed by Brecht, thanks!
Baking job is owned by scene but only need to update image space
contexts. This leads to job progress bar stuck in node editor.
Made it so node editor does not display baking jobs now.
Seems to be a regression when Campbell was working on T24009.
Bind operator exec handles curves nicely, could not see reason
why to disable this with a poll function.
From quick tests everything seems to be just fine.
Revert "Code cleanup: remove unused block from multire baker"
This reverts commit 63b01f6bee.
Multires displacement baker in fact uses level 0 for the
original subdivided mesh. Missed this when was making an
original commit.
Discard the render database when stopping the rendered preview,
otherwise starting it again may cause a segmentation fault because
undo/redo with non-rendered preview corrupts the database.
Unless I'm missing something here (probably with regards to parenting),
it makes more sense that constraint results are considered here as well
(for example, if Limit Scale constraints get applied on the object),
as this allows for greater flexibility when creating setups with this.
* Update readme for 2.70 (content + links), also updates for new tracker/git.
* Fix some links to the new blender.org website
* Release logs now point directly to the wiki, I don't see a reason to point to the website, just to redirect to the wiki after all.
the end, any reference of the object pointed by it will now lead to a RuntimeError
instead of returning None, with the aim of forcing Python API users to check the
end of iteration rather than implicitly indicating the error condition.
Acknowledgement to flokkievids for API discussions in the BlenderArtists.org
Freestyle for Blender thread.
This adds vertex snapping capabilities for curves. Snaps to all control points
of other objects, and visible + selected control points and handles in curve
edit mode.
Reviewed By: brecht
Differential Revision: http://developer.blender.org/D3
Current redraw options also did an unnecessary normal recalculation on
updated nodes.
Also, for the box and lasso mask only push an undo node if any vertex
has actually been influenced.
In the case that there are two beveled edges with one unbeveled
one in between, and the widths don't allow them to magically
line up, it is better to slide along unbeveled edge.
Sometimes bevel widths are uneven (this was the case before)
and it is a followup TODO to do a width cleanup pass afterwards
to even the edges up as much as possible.
Summary:
Old idea with changes since previous release tag
didn't work good enough. In most of the cases tag
was done in a branch hence not actually reachable
from the master branch.
Now change since release is gone, and date of
the latest commit is used instead.
The date is displayed in format YYYY-MM-DD HH:mm
in the splash.
New bpy.app fields:
- build_commit_timestamp is an unix timestamp of
the commit blender was build from.
- build_commit_date is a date of that commit.
- build_commit_time is a time of that commit.
Reviewers: campbellbarton
Differential Revision: http://developer.blender.org/D5
For materials using AO pass, this makes the material preview and the GLSL
preview more accurate, but shouldn't affect final rendering in most cases
because we usually enable AO when using the AO pass in node tree.
Thanks to Brecht for code review.
Another Evil Typo (r) one, you could add much more than the 8 allowed VCol layers!
Note: added some (warning-only) checks in mesh validate functions, but we still have a big issue with new cdlayer merge function, which could generate more than 8 layers of UVs or VCol... Don't know yet how to handle this situation. :(
